Cambridge Associates released a report this week with some compelling data about the venture capital (VC) market.
For those unfamiliar with Cambridge Associates, the firm advises large pension funds, endowments, and family offices on their allocations to various money managers. Founded in the early 1970’s, Cambridge Associates was the first firm to focus heavily on VC. At the time, commitments to VC funds were puny — only a few hundred million in aggregate each year.
The firm’s timing was impeccable with respect to two regulatory developments, as chronicled by Paul Gompers in his 1994 essay “The Rise and Fall of Venture Capital”:
The first was the 1978 Revenue Act, which decreased the capital gains tax from 49.5% to 28%. The second was the change in ERISA’s “prudent man” rule in 1979, which explicitly allowed pension funds to invest in venture capital….Prior to that date, the Employee Retirement Security Act of 1974 prohibited pension funds from investing substantial amounts of money in venture capital or other high-risk asset classes.
Annual commitments to VC funds subsequently rose by 30x to over $6 billion by 1983, with pension funds accounting for nearly half of inflows. Moreover, returns went up with all this additional capital. Median fund returns exceeded 30% at the decade’s peak in 1982, driven by a number of IPOs never-before-seen in technology.
The conventional wisdom on a 30x inflow of capital to an asset class is that it drives returns down due to simple laws of supply and demand. Clearly, the opposite occurred. What naysayers neglected to realize in the 1980’s was how much additional tech could be productized with the raw materials left over from the semiconductor boom of the 1960’s and 1970’s. Pension funds essentially created a market by funding VCs, who in turn funded a generation of remarkable founders.
This historical context is relevant for the aforementioned report. It argues that institutional investors should allocate an even greater portion of their portfolio to venture capital. I’ve summarized a few of the report’s observations below.
First, top decile institutional investors in 2019 have an average allocation to VC of 15%, nearly double their allocation in 2009.
The dispersion in returns is likely driven by superior access for top institutional investors into top VC firms. In other words, long-term devotion to the asset class tends to reward those who stick with it disproportionately.
Second, while the industry’s top VC firms tend to persist, new and emerging VCs also drive significant returns for each vintage. From 2009-2016, an average of 7 of the top 10 performing US VC funds have been new (Fund I, II) or developing (III, IV). An important caveat is that the TVPI metric tends to favor smaller funds, where higher multiples are more common, and new funds tend to be smaller. Yet, those small funds often get bigger and become new franchises over time. Hence, limited partners have myriad opportunities to get into the next great fund, if they’re willing to take risk on a new manager.
Third, venture capital is far less risky than it used to be. Loss ratios after the dot-com bubble were “only” 20% on average, while impairment ratios were 41% over the same period. These metrics exceeded 52% and 66%, respectively, from 1991-2001. Cambridge Associates posits that VC funds present lower risk due to greater portfolio diversification, as well as shorter times and capital exposed between company development milestones.
Lastly, the private equity markets, including VC, are “eating” public equity. The number of publicly listed companies in the U.S. has fallen by half over the past two decades to 4,336. Over this period, a late stage VC market has risen dramatically. 8,352 unrealized and partially realized companies were funded cumulatively by 2019. In aggregate, the market cap of the latter now more than 10% of the former. Globally, Cambridge Associates counts $340 billion of Net Asset Value (NAV) of VC-backed companies — a paltry 0.5% of the $85 trillion of global public equities.
After 40+ years of broad access to institutional capital, this data suggests that VC as an asset class could grow even larger in the future. Moreover, if we use the period of the 1980’s as a comparison, that growth does not necessarily imply lower returns. Just as companies in the 1980’s built their products on top of a mature semiconductor industry, companies in the 2020’s will benefit from prior innovations in the web, cloud computing, mobile, machine learning, wireless broadband, and open source software.
As the surface area of technology broadens and builds upon the work of those who came before, don’t be surprised to see VC grow even larger as an asset class in the coming decade.